The present paper reports the results of the experiments on the production of phytoalexin (PA) by the host-parasite interaction using rice plant as the host and Xanthomonas oryzae, the causal organism of rice leaf-blight, as the parasite. The experimental methods and the results were as follows. 1) Leaf-blade pieces of rice plant were sticked with capillary glass tube and mounted with drops of distilled water suspension of X. oryzae on the injured parts, and kept in moist chamber for 24∼48 hours. For the control, distilled water were mounted on the pricks of the leaf pieces (control 1) and the bacterial suspension were kept in test tube (control 2). Drops were collected after above noted hour and centrifuged for 10 minutes by 6000rpm. The supernatants were then added with the new suspension of X. oryzae, and after 24 hours, bacterial population of these solutions were compared by the method for the comparison of bacterial populations by phage technique14). The supernatant of X. oryzae supension on leaf inhibited the multiplication of X. oryzae (Table 1). The inhibitory action seems to be due to PA, but can not be said exactly, because X. oryzae showed no multiplication in the control 2, the supernatant of the suspension in test tube. It is presumed that the inhibition of the multiplication shown in the control 2 may be due to the deficiency of some nutrients in the solution. The second experiment, therefore, was carried out with the synthetic medium instead of distilled water: the bacterial suspension made with the synthetic medium was mounted on the leaf pricks. For the control, the same suspension was kept in test tube (control 2) and the medium only was mounted on the leaf pricks (cnotrol 1). In the present experiment, X. oryzae showed a good multiplication in the controls, and therefore the inhibition of the multiplication of X. oryzae showen by the supernatant of the bacterial suspension on the needle pricks was considered to be caused by PA (Table 2). The results mentioned above show that PA is produced by the results of the interaction between host plant and the bacterial pathogen as well as in the case of host and fungus8). 2) The antibiotic action of the PA-solution mentioned above was compared with the one produced by the interaction of rice plant and Piricularia oryzae, the causal fungus of rice-blast. The technique adopted in the case of P. oryzae was similar to the previous paper8). The results given in Table 3 and 4 show that each of these PA-solutions inhibited both the multiplication of X. oryzae and the germination of conidia of P. oryzae. It seems therefore that each of these two PA-solutions have no specificity on its antibiotic action. 3) Experiments were carried out to compare the multiplications of X. oryzae in the drops placed on the needle pricks of healthy leaves with that of ether-narcotized ones. The bacterial suspensions were mounted on the pricks. The suspension was collected after 24 hours and the bacterial populations with in unit volumes of these drops were compared by means of the bacteriophage technique. The results show that the multiplication of X. oryzae was inhibited in the case of the healthy leaves (Table 5). From the results, it is presumed that the main inhibiting factor may be due to PA, because the factor seems to exsist in the drops from healthy leaves and no PA action seems to be present in the case of the narcotized leaves.